Assessment of COPD Overdiagnosis as an Important Factor in COPD Readmission

Pandey S; Ojha S

Research Article

Austin J Pulm Respir Med. 2021; 8(2): 1072.

Assessment of COPD Overdiagnosis as an Important Factor in COPD Readmission

Pandey S¹ and Ojha S²*

¹Hospitalist Medicine, Elkhart General Hospital, Beacon Health System, Elkhart, IN, USA

²Pulmonary Medicine, Bellin Hospital, Green Bay, WI, USA

*Corresponding author: Shishir Ojha, 744 S Webster Avenue, 7^th Floor B, Green Bay, WI, 54301, USA

Received: April 15, 2021; Accepted: May 03, 2021; Published: May 10, 2021

Abstract

Introduction: The Hospital Readmission Reduction Program (HRRP) was established in 2012 to improve health care by linking payment to the quality of hospital care. Readmission is considered a hospital care quality measure. Under the program, hospitals are penalized for Chronic Obstructive Pulmonary Disease (COPD) readmission, which incentivizes improved care to avoid financial penalties. The effect of COPD overdiagnosis on COPD readmission has not been studied.

Objective: The study aims to assess the effect of COPD overdiagnosis in outpatient and inpatient settings on hospital COPD readmissions.

Methods: We conducted a retrospective study and examined outpatient and inpatient settings for COPD overdiagnosis. In the outpatient setting, we collected all COPD referrals to our clinic and reviewed charts to determine if those patients had COPD or an alternate diagnosis after our workup. We also studied 3-year inpatient data from January 2015 to March 2018 on hospital readmissions and extracted COPD readmissions. For patients seen by a pulmonary provider in our clinic, we studied patients’ pulmonary function test/ spirometry results and charts and determined if they had a true COPD diagnosis or an overdiagnosis. We also assessed the effect of COPD overdiagnosis on inflation of COPD readmission numbers.

Results: Of patients referred to our clinic, 46% did not have COPD on our workup. Among inpatients, our results revealed that preventing COPD overdiagnosis could have reduced admissions attributable to COPD by 22.6%.

Conclusion: Correct diagnosis using the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria is an inexpensive way for hospitals to avoid readmission penalties.

Keywords: Hospital readmission; COPD readmission; COPD overdiagnosis

Abbreviations

HRRP: Hospital Readmission Reduction Program; COPD: Chronic Obstructive Pulmonary Disease; PFT: Pulmonary Function Test; AECOPD: Acute Exacerbation of Chronic Obstructive Pulmonary Disease

Introduction

The Hospital Readmission Reduction Program (HRRP) was established under the Affordable Care Act and provides financial incentives to hospitals to lower hospital readmissions [1]. HRRP defines readmission as a patient who has an unplanned readmission to the same or a different acute care hospital within 30 days of discharge [2]. Approximately 20% of Medicare patients are readmitted within 30 days of their hospital discharge [3]. For inpatient Medicare beneficiaries, hospitals receive reimbursements through the Inpatient Prospective Payment System [4]. Under HRRP, the Center for Medicare and Medicaid services was required to reduce payments to hospitals under the Inpatient Prospective Payment System program for Medicare patients for excess readmissions under certain conditions starting October 1^st, 2012 [1]. Originally, the excess readmissions were determined by calculating the excess readmission ratio by comparing a hospital’s readmissions to those of all hospitals in the US with ≥25 discharges for the penalized condition [5]. The program expanded to include Acute Exacerbation of Chronic Obstructive Pulmonary Disease (AECOPD) to the list of penalized diagnoses from fiscal year 2015 (starting October 1^st, 2014), with a current maximum penalty of 3% [2].

The program aims to improve health care by linking payment to hospital care quality [1]. From 2010 to 2015, readmission rates for 3 HRRP-targeted diagnoses (acute myocardial infarction, pneumonia, and heart failure) have decreased in Medicare, Medicaid, and privately insured patient populations [5]. Conversely, the penalties adversely affected the finances of safety-net hospitals caring for uninsured and low-income patients [6]. To reduce financial penalties, hospitals have invested in strategies to reduce readmissions; several programs, such as the Care Transitions Intervention [7], Project Re-Engineered Discharge [8], and Interventions to Reduce Acute Care Transfers [9], and certain measures, such as having a skilled nursing facility within the hospital [10], have reduced readmissions.

Approximately 10%-55% of readmissions for AECOPD are preventable [11]. Several disease-specific interventions, such as pulmonary rehabilitation [12], treatment of AECOPD at home [13], inhaler device training prior to discharge [14], tele health care [15], and early outpatient follow-up [16], have been shown to improve early readmission for AECOPD.

However, none of the prior studies has investigated correcting or addressing the practice of overdiagnosis of COPD as an intervention. No prior study has investigated the effect of overdiagnosed COPD on the financial penalties paid by hospitals because of COPD readmissions.

According to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines, COPD is diagnosed by a Pulmonary Function Test (PFT) or spirometry showing a post bronchodilator Forced Expiratory Volume in 1sec (FEV1)/Forced Vital Capacity (FVC) ratio <0.7 in patients with appropriate symptoms and a history of exposure to noxious stimuli [17]. COPD overdiagnosis is common because many health care providers are not aware of its diagnostic criteria. Based on our practice, we believe that a COPD diagnosis can become a part of a patient’s medical history under several circumstances. Dyspnea in patients with a significant smoking history is often treated as COPD by primary care providers without performing spirometry. Additionally, dyspnea is commonly diagnosed as emphysema in emergency departments or urgent care if there is any degree of hyperinflation on chest X-ray images. Once a diagnosis of COPD is recorded in a patient’s medical history, it is carried over on subsequent medical visits.

The study aims to estimate COPD overdiagnosis in outpatient and inpatient settings at our North Central Montana Hospital serving 13 counties with a population of 164,000 [18]. For the outpatient setting, we examined new patients referred to our pulmonary clinic with a prior COPD diagnosis. For the inpatient setting, we analyzed our readmission database from fiscal year 2015 onward to see if COPD overdiagnosis overestimated the COPD readmission numbers.

Methods

We studied two different aspects of COPD overdiagnosis. In lieu of an Institutional Review Board, our hospital-based risk management department approved this study and waived patient consent because of the retrospective design.

The outpatient arm of the study involved a retrospective chart review of new patients referred to us with a COPD diagnosis. We included all new patients with a referral COPD diagnosis, chronic bronchitis, COPD exacerbation, or emphysema. Three providers in the pulmonary clinic see new patient referrals. The new patient referrals originate from primary care clinics, emergency departments, urgent care, and hospital discharges from various hospitals in North Central Montana.

All of the new referrals are distributed among the three providers, and the patient’s name, referral diagnosis, and date of referral are handwritten in a book maintained by the nursing staff.

We collected the number of all new referrals made in 6 months from August 2017 to February 2018. Patients who were seen until June 2018 were included in the study; no additional data were collected from any patient seen after June 2018. The referral diagnosis was extracted by the nursing staff from the faxed progress note or clinic note from the referring physician or referral order. Limited data collection, specifically the referral diagnosis, was noted from all referrals to the clinic during the study period. For the patients who were referred to the author with a COPD diagnosis, we collected more extensive data by performing a chart review of the clinic visit and gathering demographic information, age, sex, origin of referrals, hospital vs. clinic, PFT or spirometry results, smoking history, and alternate diagnosis if they did not have a COPD diagnosis from our evaluation during the pulmonary visit(s).

In the second part of our study, we evaluated the data of patients who were readmitted to an acute care hospital ≤30 days of discharge from our hospital in a 3-year period. The data were obtained from a readmission matrix maintained electronically by a hospital readmission reduction program on an Excel (Microsoft Corp., Redmond, WA) sheet per the Affordable Care Act requirements. At study conception, the data were complete and up to date from January 2015 to March 2018. The database included all-cause readmission ≤30 days of discharge after an index admission for any cause and the six penalized diagnoses (COPD, acute myocardial infarction, postcoronary bypass grafting, pneumonia, and elective total hip arthroplasty or total knee arthroplasty) as a part of the hospital readmission reduction program. All patients >18 years old and all insurance types were included. Specifically for COPD, the COPD diagnosis used to maintain the database included patients who were admitted with a primary diagnosis of COPD or primary diagnosis of respiratory failure with a secondary diagnosis of AECOPD. Anycause admission to any hospital ≤30 days of hospital discharge from the studied hospital with above criteria were included in this database as a COPD readmission [19].

If any patients were readmitted with an admitting diagnosis of COPD, we searched for a complete PFT or spirometric examination in the inpatient or outpatient electronic medical record system. If a patient had spirometry/PFT results in any setting, further chart review of the clinical notes specifically from a pulmonary clinic provider was done in the outpatient electronic medical record system to ascertain if they met the diagnostic criteria for COPD. For the study we required an FEV1/FVC ratio of <0.7 and current or past smoking history to make a diagnosis of COPD.

Statistical analysis was performed by IBM SPPS statistics for Windows (IBM Copr., Armonk, N.Y., USA) package.

Results

From August 2017 to February 2018, 519 new referrals were received by our clinic (Figure 1). Of the 519 referrals, 269 were seen by the author and 250 by the other two providers. There were 39 different referral diagnoses of which the top 15 are listed in Table 1. COPD, dyspnea on exertion, and asthma were the top three referral diagnoses. One hundred and eighteen new COPD referrals (COPD and AECOPD) and fifteen new emphysema referrals were received. Combined, there were 133 new COPD referrals of which 74 were assigned to the author and 59 to other providers.

Figure 1: Flow chart illustrating selection of the subgroup used for analysis in the outpatient arm of the study.

    
    
    Figure 1: Flow chart illustrating selection of the subgroup used for analysis in the outpatient arm of the study.

Table 1: Outpatient referrals by diagnosis.




  
    Referral diagnosis
    Author (SO)
    Others
    Total
    Percent
  
  
    COPD
    64
    54
    118
    22.7
  
  
    Emphysema
    10
    5
    15
    2.89
  
  
    Dyspnea
    39
    32
    71
    13.68
  
  
    Asthma
    22
    43
    65
    12.52
  
  
    Lung nodule
    21
    7
    28
    5.39
  
  
    Cough
    16
    27
    43
    8.28
  
  
    Not documented
    8
    14
    22
    4.23
  
  
    Lung cancer
    11
    2
    13
    2.5
  
  
    Lung Mass
    8
    2
    10
    1.92
  
  
    Hypoxemic respiratory failure
    5
    8
    13
    2.5
  
  
    Abnormal chest CT scan
    5
    7
    12
    2.31
  
  
    Pre-op
    7
    3
    10
    1.92
  
  
    Pneumonia
    5
    4
    9
    1.73
  
  
    Amyotropic lateral sclerosis/myotonic dystrophy
    6
    2
    8
    1.54
  
  
    Pulmonary fibrosis
    4
    3
    7
    1.34



Table 1: Outpatient referrals by diagnosis.

Out of the 74 assigned, the author eventually saw 61.

Of these 61 patients, twenty-eight patients were found not to have COPD (Table 2). Of these 28 patients, 9 had never smoked a cigarette, but 1 had smoked marijuana. The other 33 patients had accurate COPD diagnoses (Table 3). Hence, 46% of the patients referred with a COPD diagnosis did not have COPD on our workup.

Table 2: Other diagnoses after work up.




  
    S.N0
    Referral Diagnosis
    Age (years)
    Fev1/FVC ratioa
    Gender
    Smoking history (pack- years)
    Final diagnosis or    impression/comments
  
  
    1
    COPD
    63
    75
    M
    25
    Unknown
  
  
    2
    COPD Exacerbation
    68
    83
    F
    38
    Asthma
  
  
    3
    COPD
    54
    79
    F
    41
    Unknown
  
  
    4
    COPD
    42
    72
    M
    Never smoker
    Lingular PNA
  
  
    5
    COPD
    63
    74
    M
    48
    Lingular PNA
  
  
    6
    Emphysema
    74
    79
    M
    Never smoker
    Chronic indolent infection causing cough,    scarring
  
  
    7
    COPD
    76
    76
    M
    20
    Unknown
  
  
    8
    COPD
    60
    74
    F
    47
    Dyspnea at rest, not with exertion
  
  
    9
    COPD
    54
    86
    F
    Marijuana
    Mild asthma, positive bronchodilator response
  
  
    10
    COPD
    59
    81
    M
    40
    Deconditioning, weight 253 pounds
  
  
    11
    Emphysema
    77
    67,73
    F
    Never smoker
    Asthma, reversible obstruction, positive    bronchodilator response for FEV1
  
  
    12
    COPD
    43
    71
    M
    26
    Asthma
  
  
    13
    COPD
    67
    65,72
    M
    Never smoker
    Asthma, reversible obstruction
  
  
    14
    Emphysema
    64
    78
    F
    45
    Normal coronaries, mild symptoms suggestive for    asthma
  
  
    15
    COPD
    63
    81
    F
    45
    Asthma based on history
  
  
    16
    COPD
    82
    73
    F
    19
    Paralyzed and elevated right hemidiaphragm
  
  
    17
    COPD
    76
    70
    M
    12
    Equivocal for COPD
  
  
    18
    COPD
    58
    75
    M
    45
    Usual interstitial pneumonitis
  
  
    19
    COPD
    59
    81
    M
    40
    Asthma
  
  
    20
    COPD
    62
    81
    M
    5
    Denied dyspnea during clinic visit
  
  
    21
    COPD
    61
    61,64
    M
    Never smoker
    Asthma since 1997
  
  
    22
    COPD
    67
    79
    F
    Never smoker
    Cough variant asthma, 24% and 450 cc increase in    FEv1.
  
  
    23
    COPD
    68
    76
    F
    9
    Acute bronchitis, resolved by the time of visit
  
  
    24
    COPD
    56
    59,73
    M
    Never smoker
    Mild intermittent asthma
  
  
    25
    COPD-asthma
    59
    72
    M
    Never smoker
    Mild intermittent asthma
  
  
    26
    COPD
    68
    64,66
    F
    35
    Dyspnea on hiking, climbing mountains
  
  
    27
    COPD
    63
    50,56
    F
    Never smoker
    Sarcoid, asthma
  
  
    28
    COPD
    66
    72
    M
    27
    Non ischemic cardiomyopathy
  
  
    aOnly post bronchodilator Fev1/FVC ratio is    provided for all the patients. If pre-bronchodilator ratio was <70 then    both pre- and post-ratios are provided.



Table 2: Other diagnoses after work up.

Table 3: COPD referrals, which were confirmed as COPD after work up.




  
    S.NO.
    Referral Diagnosis
    Gender
    Age
    Fev1/FVC ratio
    Fev1
    Smoking pack-years
    Final Diagnosis
  
  
    1
    COPD
    F
    61
    41
    3
    32
    COPD
  
  
    2
    COPD
    M
    62
    61
    3
    38
    COPD
  
  
    3
    COPD
    F
    73
    26
    4
    128
    COPD
  
  
    4
    COPD
    M
    78
    34
    3
    41
    COPD
  
  
    5
    COPD
    F
    65
    38
    4
    41
    COPD
  
  
    6
    COPD
    M
    52
    64
    2
    19
    COPD
  
  
    7
    COPD
    M
    72
    66
    2
    40
    COPD
  
  
    8
    COPD
    M
    72
    53
    3
    96
    COPD
  
  
    9
    COPD
    F
    59
    64
    3
    50
    COPD
  
  
    10
    COPD
    M
    79
    67
    1
    58
    COPD
  
  
    11
    COPD
    F
    61
    35
    4
    40
    COPD
  
  
    12
    COPD
    M
    61
    57
    2
    135
    COPD
  
  
    13
    COPD
    M
    72
    60
    2
    58
    COPD
  
  
    14
    COPD
    F
    57
    60
    2
    36
    COPD with asthma
  
  
    15
    COPD
    F
    80
    63
    2
    30
    COPD
  
  
    16
    COPD, asbestosis
    M
    80
    63
    3
    30
    COPD
  
  
    17
    COPD, pre-op clearance
    M
    79
    28
    4
    75
    COPD
  
  
    18
    COPD
    F
    62
    38
    2
    18
    COPD
  
  
    19
    COPD
    M
    57
    66
    2
    76
    COPD
  
  
    20
    COPD
    M
    71
    38
    4
    159
    COPD
  
  
    21
    COPD
    M
    62
    61
    2
    27
    COPD with asthma
  
  
    22
    COPD
    M
    80
    53
    2
    91
    COPD
  
  
    23
    Emphysema
    F
    69
    67
    2
    50
    COPD
  
  
    24
    COPD
    M
    65
    46
    3
    50
    COPD
  
  
    25
    COPD
    F
    67
    67
    2
    42
    COPD
  
  
    26
    COPD
    M
    69
    64
    2
    128
    COPD
  
  
    27
    COPD
    M
    68
    58
    2
    65
    COPD
  
  
    28
    COPD, pre-op clearance
    M
    61
    46
    2
    44
    COPD
  
  
    29
    COPD
    M
    67
    59
    2
    51
    COPD
  
  
    30
    COPD
    M
    63
    24
    4
    46
    COPD
  
  
    31
    COPD
    F
    75
    36
    3
    38
    COPD
  
  
    23
    COPD
    M
    69
    42
    3
    48
    COPD
  
  
    33
    COPD
    M
    76
    65
    3
    44
    COPD with asthma



Table 3: COPD referrals, which were confirmed as COPD after work up.

The patients with a final diagnosis of COPD were older, more likely to be smokers, and had more pack-years of smoking history than those with an alternate diagnosis. There was no difference in sex between the two groups (Table 4).

Table 4: Patient characteristics by confirmed COPD versus other diagnoses.




  
    
    Confirmed COPD (n=33)
    Other diagnosis (n=28)
    testa
    P-value
  
  
    Gender
    
    
    
    
  
  
    Male n (%)
    22 (66.7)
    16 ( 57.1)
    χ2
    0.444
  
  
    Female
    11
    12
    
    
  
  
    Age M (SD)
    68.0 (7.7)
    63.3 (9.2)β
    
    0.033
  
  
    Smoking history n (%)
    33 (100)
    19 (70.3)β
    p
    <0.001
  
  
    Smoking pack-years M-(SD)
    58.3 (35.0)
    30.9 (13.8)
    U
    0.001
  
  
    Fev1/FVC ratio M (SD)
    51.8 (13.8)
    74.7 (6.2)
    t
    <0.001
  
  
    aTest χ2 = Chi-Square; t = independent t-test; p = Fisher’s    exact test; U = Mann-Whitney U-test

      βOne patient who only smoked marijuana was    excluded from the smoking comparisons



Table 4: Patient characteristics by confirmed COPD versus other diagnoses.

The 30-day readmission matrix showed that, during the 3-year period from January 2015 to March 2018, there were a total of 4583 admissions (Figure 2) and 444 30-day all-cause readmissions for the 6 penalized diagnoses (Table 5).

Figure 2: Flow chart illustrating selection of subgroup selection for analysis in the inpatient arm of the study. ^aPrimary diagnosis of COPD or primary diagnosis of respiratory failure with a secondary diagnosis of acute exacerbation of chronic obstructive pulmonary disease.

    
    
    Figure 2: Flow chart illustrating selection of subgroup selection for analysis in the inpatient arm of the study.
^aPrimary diagnosis of COPD or primary diagnosis of respiratory failure with a secondary diagnosis of acute exacerbation of chronic obstructive pulmonary disease.

Table 5: Penalized readmissions by diagnosis (inpatient).




  
    S.NO
    Diagnosis
    Total
    Readmits
    Percentage readmits
  
  
    1
    PNAa
    1157
    134
    11.5
  
  
    2
    TKA/THAβ
    1135
    29
    2.55
  
  
    3
    Heart failure
    716
    118
    16.4
  
  
    4
    COPD
    615
    88
    14.3
  
  
    
    COPD Medicare*
    365
    53
    14.5
  
  
    5
    CABG†
    236
    25
    10.5
  
  
    6
    AMI‡
    724
    50
    6.9
  
  
    Total
    
    4583
    444
    9.68
  
  
    aPNA,    Pneumonia.

      βTKA,    total knee arthroplasty, THA, total hip arthroplasty.

      *Medicare    data are a subgroup of S.NO 4 COPD.

      †CABG,    coronary artery bypass grafting.

      ‡AMI, acute myocardial infarction.



Table 5: Penalized readmissions by diagnosis (inpatient).

There were total 615 COPD admissions, including 88 readmission cases involving 75 patients (Table 6). Of these readmission cases, only 38 cases had either COPD as the primary diagnosis or respiratory failure as primary with AECOPD as a secondary diagnosis during readmission. Thus, 50 readmission cases were due to causes other than COPD (Figure 2).

Table 6: Readmission episodes by gender and recurrence.a




  
    
    No. of patientsa 
    Total episodes
  
  
    Males (n=31)
     
     
  
  
    1 re-admission
    28
    28
  
  
    2 re-admissions
    1
    2
  
  
    3 re-admissions
    2
    6
  
  
    Females (n=44)
    
    
  
  
    1 re-admission
    38
    38
  
  
    2 re-admissions
    4
    8
  
  
    3 re-admissions
    2
    6
  
  
    Total
    75
    88
  
  
    aRecurrence: Repeat readmission episode for the    same patient.



Table 6: Readmission episodes by gender and recurrence.a

The mean age was 70 years for males and 67 years for females. Forty-four patients had undergone PFT or spirometry and were seen in the network affiliated with the hospital by a board-certified pulmonologist or an experienced physician assistant. Thirty-one patients were not referred to a pulmonary clinic in the network and PFT/spirometry were not available for those patients. Out of these 44 patients, 34 patients had confirmed COPD based on their spirometry and smoking history obtained on chart review. Ten patients did not have COPD. The FEV1/FVC ratio, FEV1, and COPD stage are presented in Table 7.

Table 7: Characteristics of inpatients with confirmed COPD versus other diagnosis.




  
    
    Other diagnosis (n = 10)
    Confirmed COPD (n = 34)
    testa
    p value
  
  
    Gender
    
    
    
    
  
  
    Male n (%)
    4 (38.2)
    13 (40.0)
    p
    1
  
  
    Female n (%)
    6
    21
    
    
  
  
    Age in years M (SD)
    68.6(20.7)
    70.9(10.5)
    t
    0.637
  
  
    Fev1/FVC ratio-M (SD)
    75.2(6.9)
    45.1 (11.6)
    t
    <0.001
  
  
    Fev1 absolute(mls)-M (SD)
    1818.0(695.3)
    948.5(376.1)
    U
    0.001
  
  
    FEV1 percent predicted
    69.8(20.7)
    40.4(15.3)
    t
    <0.001
  
  
    COPD stage
    
    
    
    
  
  
    I
    
    1
    
    
  
  
    2
    
    8
    
    
  
  
    3
    
    18
    
    
  
  
    4
    
    7
    
    
  
  
    Other Diagnosis
    Restrictive, 2
    
    
    
  
  
    
    Asthma, 2
    
    
    
  
  
    
    ILD, 1
    
    
    
  
  
    
    Cough, 1
    
    
    
  
  
    
    Diastolic dysfunction, 1
    
    
    
  
  
    
    Tracheal-stenosis, 1
    
    
    
  
  
    
    Cystic fibrosis, 1
    
    
    
  
  
    
    Trapped lung, 1
    
    
    
  
  
    ap = Fisher’s exact test;    t = independent t-test; U = Mann–Whitney U test.



Table 7: Characteristics of inpatients with confirmed COPD versus other diagnosis.

There were no differences in the age and sex distribution between the groups, but the FEV1/FVC ratio, absolute FEV1 in milliliters, and predicted FEV1 as a percentage were significantly lower in the patients with COPD than in those without.

There were 53 readmissions for these 44 patients. There were 41 readmissions for the 34 patients confirmed with COPD; 2 patients had 3 readmissions, and 3 had 2 readmissions. Ten patients who did not have COPD accounted for 12 readmissions because 1 patient was readmitted thrice. Hence, there were 12 readmissions out of 53 over 3 years, which was attributed to COPD overdiagnosis. Correct diagnosis could have led to a 22.6% decrease in admissions attributable to COPD. Of these 10 patients, 7 had Medicare as their primary insurance, 2 had Medicaid, and 1 had Blue Cross of Montana.

Discussion

In the US, there are 1.5 million emergency department visits and 725,000 hospital admissions for COPD each year, with a health care cost of about $60 billion [20,21]. Among Medicare patients, 17.6% are readmitted ≤30 days of discharge, which increases to approximately 20% for COPD patients. The HRRP reduces payments to hospitals with excess COPD readmissions, with a current reduction cap of 3%. The program goal is to improve health care by linking payment to hospital care quality [1]. Program critics argue that the program does not account for differences in the socioeconomic status between hospitals, penalties are too large, and hospitals cannot improve patients’ compliance [22]. The hospitals that must pay HRRP penalties serve vulnerable populations with many low-income patients [23]. To address these concerns and achieve more equitable distribution of penalties, the Center for Medicare and Medicaid services changed the methodology for payment reduction [2,24]. Starting in 2019, the hospital performance is assessed relative to that of other hospitals with similar proportions of patients eligible for Medicare and fullbenefit Medicaid [2]. Currently, excess readmissions are measured by the excess readmission ratio [5].

Despite the penalties for COPD readmission, hospitals may be incurring those penalties for patients admitted with other cardiorespiratory diseases because of COPD overdiagnosis. As found in our study, COPD overdiagnosis is prevalent in both new referrals and readmissions. This self-caused penalty is caused by physicians not using validated COPD criteria in initial assessments. In our experience, many non-pulmonary physicians “diagnose” COPD according to the patient’s history of cough, breathlessness, and smoking without using spirometric data. This approach is in striking contrast with the diagnosis of congestive heart failure, commonly by echocardiography [25]. Approximately half of primary care physicians are unaware of the COPD GOLD guidelines and use clinical criteria to diagnose COPD because they think that spirometry is unnecessary [26,27].

The reason for the underuse of spirometry to diagnose the etiology of dyspnea may be a lack of proper training [27]. Many physicians do not know the indications of spirometry or how to order one [28]. We believe that underuse could also be because misdiagnosis is not directly life threatening. For immediate life-threatening conditions, such as pulmonary embolism or myocardial infarction, the issue is over-testing [29,30] and not necessarily overdiagnosis by clinicians without proper testing.

Most clinicians believe that patients are too weak to undergo a hospital PFT; there may be a difference in the numbers obtained during an exacerbation and in the outpatient setting. However, several recent studies found that a hospital spirometry prior to discharge was not different from one during clinical stability and inpatient spirometry provided good accuracy to predict outpatient airflow obstruction, although it overestimated severity [31,32]. Even if inpatient-setting spirometry may overdiagnose COPD in a small number of patients [31], the benefit in eliminating many erroneous COPD diagnoses will help allocate hospital resources for investigating the true diagnosis.

Another reason for overdiagnosis could be that COPD is more common than many other pulmonary conditions, such as interstitial lung diseases [33]. Providing a prescription for an inhaler to treat “COPD” is much quicker [34] than investigating comorbidities, such as obesity or cardiac issues, or testing for interstitial lung diseases in a busy practice. Hence, COPD is used very liberally as a catchall for all respiratory ailments.

The impact of COPD overdiagnosis is not limited to incurring avoidable penalties for readmissions. Overtreatment of COPD has adverse health consequences in terms of drug-related adverse effects and potential financial costs [35]. Additionally, treatment of the true diagnosis is delayed. Lack of proper treatment between admissions may itself contribute to readmission. In the current insurance milieu, documenting spirometry data in COPD exacerbations should be a part of history taking for patients admitted for COPD exacerbation. If no PFT data is available, spirometry should perhaps be a part of inpatient COPD management.

All other efforts to improve COPD management during the index hospitalization and post-discharge periods may not reduce “COPD readmission”. As observed in our study, in 50 of 88 COPD readmission episodes the admitting diagnosis during the readmission episode, was not COPD (Figure 2) and included conditions unrelated to respiratory diseases, like cellulitis and urinary tract infection. However, since the admitting diagnosis during the index admission was COPD, those readmission episodes were still subject to financial penalties. Hence, better control of COPD alone during the index admission or post hospital discharge may prove futile in preventing a readmission. A more effective strategy would be to avoid COPD overdiagnosis at the index admission or in the outpatient setting. In our study, COPD overdiagnosis resulted in 29% excess readmissions for COPD. The national COPD readmission rate is 19.5%. The latest COPD readmission rate for the hospital 14.5%, is no different than the HRRP’s “expected readmission rate” as the mean national readmission rate [36], and it has stayed below the expected rate since the penalty inception despite overdiagnosis.

Eliminating the readmission load due to COPD overdiagnosis may not be difficult. Many hospitals have established a COPD readmission team led by a pulmonologist. A COPD navigator who is a either a nurse, nurse practitioner, respiratory therapist, or technician from the PFT lab can be assigned to screen all inpatient charts of patients admitted with a diagnosis. If the patient has had a prior PFT or spirometry with a post-bronchodilator FEV1/FVC ratio =70, the attending physician should be made aware, which may cost fewer resources than the number some hospitals assign for pulmonary consultations with every COPD patient. However, this measure will not eliminate all patients without COPD because some of the PFT/ spirometry done at other hospitals or out-of-network clinics may not be available.

Absence of COPD essentially rules out a COPD exacerbation diagnosis. However, a correct COPD diagnosis alone does not necessarily mean that a dyspneic, hypoxemic, or hypercarbic patient is in COPD exacerbation because other comorbidities such as congestive heart failure and untreated obstructive sleep apnea are frequently present [37], and some may cause a hospital admission in a COPD patient without causing COPD exacerbation [38]. Our study had several strengths. We evaluated two different settings of COPD overdiagnosis, outpatient referrals to the pulmonary clinic and inpatient readmissions, and found that COPD was overdiagnosed in both settings. We found that one single intervention, correcting the diagnosis, could result in a 22.6% reduction in readmissions attributable to COPD.

The study limitations were its retrospective nature that relied on chart review, which limited the sample size because PFT/spirometry was not available on all the charts in the study’s inpatient arm. Only in 53 readmissions (44 patients) out of 88 could we determine if patients had a true COPD diagnosis or were overdiagnosed.

However, only a retrospective study design is more suited to uncover the gravity of “COPD overdiagnosis” and its true effect on inflation of readmission numbers.

The study was conducted at a community hospital in North Central Montana. The age-adjusted prevalence rate of COPD in Montana of 5.4% is similar to the national average of 5.9% [39]. Despite the similar rates, demographics for COPD patients are not representative of a typical US population because the survey that provided this prevalence rate did not include data for minorities because of the low numbers of African American and Hispanic responders [38]. Additionally, the most current report from the Center for Medicare and Medicaid services shows that none of the Montana Hospitals was penalized for COPD readmission [36], whereas 83% of US hospitals received some penalties for the six penalized diagnoses [40], the readmission rate across all-cause readmission is among the six lowest of states [41]. In terms of avoiding penalties, Montana has fared better than all but one or two states in the past 5 years [42,43]. These factors raise the question of whether the findings are applicable to other states.

Conclusion

In the HRRP era in which the COPD readmission penalties are at their maximal rate, hospitals need to strongly consider implementing the GOLD guidelines and inpatient spirometries to ensure a correct diagnosis of the underlying etiology of dyspnea and minimize selfcaused penalties.

Acknowledgment

The authors would like to thank Enago (www.enago.com) for the English language review.

Statistical analyses for Tables 4 and 7 were provided by an independent statistical consultant Dr. Linda Deacon.

They were compensated directly by the authors.

Author Contributions

SO takes full responsibility for the manuscript content, data integrity, and data analysis accuracy.

SP contributed substantially to the study design, data entry, and writing of the manuscript.

References

Download PDF

Citation: Pandey S and Ojha S. Assessment of COPD Overdiagnosis as an Important Factor in COPD Readmission. Austin J Pulm Respir Med. 2021; 8(2): 1072.

Home

Journal Scope

Editorial Board

Instruction for Authors

Submit Your Article